Cognitive impairment in different MS subtypes and clinically isolated syndromes

Cognitive impairment in different MS subtypes and clinically isolated syndromes

Journal of the Neurological Sciences 267 (2008) 100 – 106 www.elsevier.com/locate/jns Cognitive impairment in different MS subtypes and clinically is...

147KB Sizes 0 Downloads 63 Views

Journal of the Neurological Sciences 267 (2008) 100 – 106 www.elsevier.com/locate/jns

Cognitive impairment in different MS subtypes and clinically isolated syndromes Constantin Potagas, Erasmia Giogkaraki ⁎, Georgios Koutsis, Dimitrios Mandellos, Erifylli Tsirempolou, Constantin Sfagos, Demetris Vassilopoulos Eginition Hospital, Department of Neurology, School of Medicine, University of Athens, Greece Received 4 June 2007; received in revised form 28 September 2007; accepted 4 October 2007 Available online 13 November 2007

Abstract Objective: To investigate the pattern of cognitive impairment in patients with relapsing–remitting (RR), secondary progressive (SP), primary progressive (PP) multiple sclerosis, and patients with clinically isolated syndrome (CIS) suggestive of MS, relative to control participants in the Greek population. Methods: RR patients (N = 75), SP patients (N = 29), PP patients (N = 23), CIS patients (N = 33), and healthy control participants (N = 43) were assessed by the Brief Repeatable Battery of Neuropsychological Tests (BRBN). Results: The overall prevalence of cognitive dysfunction in our patients was 52.8% with CIS patients excluded and 47.5% with CIS patients included. All MS patients differed significantly from controls in all BRBN measures. Similar was the pattern of cognitive dysfunction in patients with CIS suggestive of MS, although verbal learning/memory capacity (as measured by the Selective Reminding Test) remained relatively spared. The comparisons between patient groups revealed some differences in the performance mainly in favor of CIS and RRMS patients. These differences largely disappeared after controlling for physical disability (EDSS). Conclusion: All MS subtypes patients exhibit a pattern of cognitive impairment running across the studied cognitive domains. The pattern of cognitive dysfunction in patients with CIS is similar with relative sparing of verbal learning. © 2007 Elsevier B.V. All rights reserved. Keywords: Brief Repeatable Battery; Cognitive impairment; Speed processing deficit; Multiple sclerosis subtypes; Clinically isolated syndromes suggestive of multiple sclerosis

1. Introduction Cognitive impairment is present in 40–65% of patients with multiple sclerosis (MS), encompassing all disease stages and types of clinical course. [1–3] These figures usually do not include patients with clinically isolated syndrome (CIS), though cognitive dysfunction has also been demonstrated in patients with CIS [4]. MS-related cognitive dysfunction is characterized by prominent involvement of recent memory, sustained attention, information processing speed, and executive functions [5]. ⁎ Corresponding author. 22, Rostand Street, 111-41, Athens, Greece. Tel.: +30 6944835744; fax: +30 2102925824. E-mail address: [email protected] (E. Giogkaraki). 0022-510X/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2007.10.002

The evidence at present suggests that relapsing–remitting (RR) patients perform better than primary progressive (PP) or secondary progressive (SP) patients on the Brief Repeatable Neuropsychological Battery (BRΒΝ) [2,6] and on several other cognitive tasks. [7] Comparative data on the performance of PP and SP patients have demonstrated more severe cognitive deficits in SP patients [8,9]. Other studies have shown significant differences in particular cognitive domains between all three subtypes (RR, PP, SP) [2,10], suggesting heterogeneity and distinct cognitive profiles depending on disease course [3]. To the best of our knowledge no single study has compared cognitive dysfunction between RR, PP, SP and CIS patients. The present study investigated differences in cognitive profile, as assessed by BRBN [6,11], between different MS

C. Potagas et al. / Journal of the Neurological Sciences 267 (2008) 100–106

subtypes (RR, PP, SP), and CIS patients and further compared performance to that of controls in the Greek population. It represents the first systematic study of cognitive function in Greek MS patients, allowing an estimation of the prevalence of cognitive impairment in the different subtypes of MS. 2. Methods 2.1. Participants One hundred sixty consecutive patients with MS under regular follow-up care at the Clinic for demyelinating diseases, Department of Neurology, University of Athens were studied. McDonald's criteria [12] were used for inclusion. Patients with acute relapse during the preceding month, severe visual or upper limb involvement interfering with neuropsychological testing, major psychiatric illness, other neurological disease, learning disability, non Greek origin or insufficient command of the Greek language were excluded from the study. Patients were classified as RRMS (N = 75), SPMS (N = 29), PPMS (N = 23), or CIS suggestive of MS (N = 33). A detailed neurological examination was obtained for all patients. Physical disability was scored using the Expanded Disability Status Scale (EDSS) [13]. In addition, 43 Greek control participants were recruited from the community, so as to obtain a sample with demographic characteristics as close as possible to our patients' sample. Their medical history was obtained by an interview preceding assessment. The volunteers were excluded in case of learning disabilities or any psychiatric or other neurological disorders, traumatic brain injury, cardiovascular illness and drug or alcohol abuse. They were all native Greek speakers and had normal visual acuity. All participants gave informed consent to participate in this study, which was approved by the Ethics Committee of the Hospital. Table 1 shows age, education, and gender distribution for patients and controls, as well as EDSS, Beck Depression Inventory [14] scores and disease duration for MS patients. 2.2. Neuropsychological assessment Neuropsychological assessment was performed with the Brief Repeatable Battery of Neuropsychological Tests

101

(BRBN) [6,11] adapted in the Greek language. The Beck Depression Inventory (BDΙ) [14] was also administered to all patients, in order to evaluate their mood. The Bushke Verbal Selective Reminding Test (SRT; 6trial version) [15] is a measure of verbal learning/memory of a 12-word list. The Long-Term Storage (SRTL) score represents the sum of words recalled on two consecutive trials without reminding. The Consistent Long-Term Retrieval (SRTC) score is the sum of words recalled on all subsequent trials without reminding. The Delayed Recall (SRTD) score is the number of words recalled after a delay of 15 min. The 10/36 Spatial Recall Test [11] measures visuospatial learning and memory. It requires participants to recall the placement of 10 checkers that are randomly placed on a 6 × 6 checkerboard. Two scores are recorded; one is the sum of correct responses in the three immediate recall trials (SPARTi), and the second is the delayed recall after 15 min (SPARTd). The Symbol Digit Modalities Test (SDMT: oral version) [16] examines speed of visual information processing, complex visual scanning, and sustained attention. Participants have to verbally substitute meaningless symbols by the corresponding number. The score is the number of correct substitutions in 90 s. The Paced Auditory Serial Addition Test (PASAT) [17] requires mental calculation, working memory and interference suppression, concentration and information processing speed. Participants are instructed to add 60 pairs of digits, such that each number is added to the one immediately preceding it, and report the outcome verbally. The digits are presented by audiotape, first at a rate of 3 s per digit (PASAT3), then, in a second trial, at a rate of 2 s per digit (PASAT2). Scores are the sums of correct responses for the 3- and 2-seconds forms of the task. The Word List Generation (WLG) [6] is a semantic verbal fluency test evaluating the spontaneous production of names of a given category (fruits and vegetables) within 90 s. The score is the number of correct words. We classified as cognitively impaired through examination with the present screening battery (BRBN), patients who failed on at least 33% (3/9) of the included measures. [18] We considered that patients had failed a particular test if they scored below the 5th percentile for controls. The frequency

Table 1 Demographic and clinical characteristics of patients and controls

N Age (years) Education (years) Gender, M/F, % EDSS Duration (years) Beck

Controls

RRMS

SPMS

PPMS

CIS

Significant differences

43 36.2 (11.3) 14.1 (3.7) 41.9/ 58.1

75 34.3 (8.9) 14.2 (2.9) 32.0/68.0 1.9 (1.6) 6.2 (4.9) 10.8 (8.6)

29 42.0 (8.5) 14.1 (2.7) 44.8/55.2 5.6 (1.3) 15.3 (7.9) 18.9 (9.5)

23 42.8 (9.9) 12.8 (3.0) 43.5/56.5 3.7 (1.6) 4.7 (5.3) 13.3 (9.0)

33 34.7 (8.7) 13.5 (3.0) 45.5/54.5 1.5 (1.2) 1.0 (1.5) 8.2 (9.1)

RR b PP, SP; CIS b SP, PP ns ns CIS, RR b SP, PP; PP b SP SP N CIS, RR, PP; RR N CIS SP N CIS, RR

Values are mean (SD); one-way ANOVA for age, disease duration, age of onset, education (p b 0,05), Mann–Whitney U test for EDSS, Pearson χ2 for gender; RRMS = relapsing–remitting; PPMS = primary progressive; SPMS = secondary progressive; CIS = clinically isolated syndrome.

102

C. Potagas et al. / Journal of the Neurological Sciences 267 (2008) 100–106

Table 2 Pearson correlation coefficients between measures of BRBN, age and education in control participants

SRTC SRTD SPARTi SPARTd SDMT PASAT3 PASAT2 WLG Age Education

SRTL

SRTC

.92⁎⁎ .75⁎⁎ .36⁎ .19 .39⁎ .34⁎ .32⁎ .12 − .43⁎⁎ .21

1

.68⁎⁎ .32⁎ .11 .42⁎⁎ .27 .27 .13 − .41⁎⁎ .22

SRTD

SPARTi

SPARTd

1 .33⁎ .16 .37⁎ .19 .20 .26 − .41⁎⁎ .20

1 .78⁎⁎ .54⁎⁎ .49⁎⁎ .45⁎ .12 − .65⁎⁎ .23

1 .35⁎ .43⁎ .33⁎ .01 −.45⁎⁎ .04

SDMT

PASAT3

PASAT2

WLG

Age

Education

1 .64⁎⁎ − .02 − .41⁎⁎ .04

1 −.06 −.34⁎ .34⁎

1 −.08 .02

1 − .29

1

1

.39⁎ .49⁎⁎ .34⁎ − .41⁎⁎ .48⁎

⁎Correlation is significant at the .05 level (2-tailed); ⁎⁎ Correlation is significant at the .01 level (2-tailed); SRT = Selective Reminding Test: L = Long-Term Storage; C = Consistent Long-Term Storage; D = Total Delay Recall; SPART = 10/36 Spatial Recall: i = immediate recall; d = delayed recall; SDMT = Symbol Digit Modality Task; PASAT = Paced Auditory Serial Addition Task (3 = 3 s/digit; 2 = 2 s/digit); WLG = Word List Generation.

of cognitive dysfunction in each measure in different MS subtypes was defined as the percent of patients classified as impaired in one task minus the percent of control participants misclassified as impaired for the same task [6].

interpreted as being small (d = 0.2), medium (d = 0.5), or large (d N = 0.8) [20].

2.3. Statistical analyses

3.1. Demographic and clinical data

Differences between groups on clinical and demographic characteristics were analyzed by means of one-way ANOVA (age, education, disease duration, BDI), independent-samples Mann–Whitney U test for rank data (EDSS) and Pearson χ2 tests (gender distribution). Pearson correlations (for age, education, disease duration, and BDI) and Spearman rank order correlations (for EDSS; bivariate, two-tailed) were calculated to examine the relations between clinical and demographic characteristics and task performance. The relation between task performance and gender was investigated by means of independent-samples Student t-tests (two-tailed). Two multivariate procedures (SPSS 13.0; Chicago IL), with all BRBN scores as dependent variables, were performed to examine differences between controls, CIS, and definite MS patients, and between MS subtypes. Age was included as covariate in both procedures because differences in age between groups of patients were significant and because age correlated negatively with most measures. Gender was also a covariate in both procedures, as a gender impact was observed in most BRBN measures in MS patients. Age and gender were significant covariates for the multivariate analysis (Wilks' λ = . 83, F(9,188) = 4.14, p b . 001 for age and Wilks' λ = . 84, F(9,188) = 3.83, p b . 001 for gender). Education was not a significant covariate (Wilks' λ = . 92, F(9,188) = 1.79, p = . 07). Paired comparisons were performed on the age and gender covariate adjusted scores [19]. Disease duration and EDSS were not included simultaneously as covariates in the multivariate comparison of performances of MS groups, to prevent overcorrection [2]. As a measure of the effect sizes, we calculated the Cohen d, which indicates the magnitude of mean differences (using the estimated marginal means) in SD units. Effect sizes can be

RRMS and CIS patients were younger than PPMS and SPMS patients. There were no significant gender distribution differences or differences in years of education between the groups. RRMS and CIS patients had lower EDSS scores than SPMS and PPMS patients and furthermore the PPMS subgroup had lower EDSS scores than the SPMS subgroup. Disease duration was longer and BDI score was higher for the SPMS group, compared to the other groups. Duration was also longer for RRMS patients compared to CIS patients (Table 1).

3. Results

3.2. The control group In the control sample, age influenced negatively all measures (moderate correlation for SPARTi and low correlations Table 3 Numbers (%) of patients impaired (b5 percentile) in each measure

SRTL SRTC SRTD SPARTi SPARTd SDMT PASAT3 PASAT2 WLG

RR (N = 75)

PP (N = 23)

SP (N = 29)

CIS (N = 33)

16 (21.3) 17 (22.0) 14 (19.0) 10 (12.7) 21 (28.0) 21 (27.7) 17 (22.0) 17 (22.0) 28 (37.3)

8 (34.8) 8 (34.5) 4 (19.4) 7 (30.1) 9 (30.1) 14 (62.6) 10 (43.2) 8 (34.5) 13 (56.5)

16 12 13 9 16 22 14 13 13

4 (12.1) 2 (7.5) 4 (12.8) 4 (12.1) 9 (27.3) 5 (14.2) 6 (19.6) 5 (16.6) 11 (33.3)

(55.2) (40.2) (46.0) (29.8) (55.2) (75.8) (47.1) (43.6) (44.8)

SRT = Selective Reminding Test: L = Long-Term Storage; C = Consistent Long-Term Storage; D = Total Delay Recall; SPART = 10/36 Spatial Recall: i = immediate recall; d = delayed recall; SDMT = Symbol Digit Modality Task; PASAT = Paced Auditory Serial Addition Task (3 = 3 s/digit; 2 = 2 s/ digit); WLG = Word List Generation; RR = relapsing–remitting; PP = primary progressive; SP = secondary progressive; CIS = clinically isolated syndrome.

C. Potagas et al. / Journal of the Neurological Sciences 267 (2008) 100–106 Table 4 Correlation coefficients between, demographic and clinical characteristics and measures of BRBN of MS patients (Pearson's correlations for age, education and disease duration and Spearman rank order correlations for EDSS) Age Education Duration EDSS BDI SRTL SRTC SRTD SPARTi SPARTd SDMT PASAT3 PASAT2 WLG

.376⁎⁎ .321⁎⁎ .015 − .343⁎⁎ − .275 − .281 − .321⁎⁎ − .312⁎⁎ − .420⁎⁎ − .236 − .214 − .155

Education

Duration

EDSS

BDI

1 1

− .149 .140 .189 .172 .192 .206 .184 .212 .127 .078

.486⁎⁎ .177 − .337⁎⁎ − .308⁎⁎ − .321⁎⁎ .247 − .273 − .373⁎⁎ − .274 − .280 − .153

1 .327⁎ − .388⁎⁎ − .414⁎⁎ − .390⁎⁎ − .265 − .287 − .510⁎⁎ − .345⁎⁎ − .253 − .259

1 − 203 − .208 − .116 − .112 − .180 − .141 − .130 − .860 .007

⁎Correlation is significant at the .05 level (2-tailed); ⁎⁎ Correlation is significant at the .01 level (2-tailed); SRT = Selective Reminding Test: L = Long-Term Storage; C = Consistent Long-Term Storage; D = Total Delay Recall; SPART = 10/36 Spatial Recall: i = immediate recall; d = delayed recall; SDMT = Symbol Digit Modality Task; PASAT = Paced Auditory Serial Addition Task (3 = 3 s/digit; 2 = 2 s/digit); WLG = Word List Generation, BDI = Beck Depression Inventory; EDSS = Expanded Disability Status Scale.

for the other measures) except WLG (Table 2). Gender influenced significantly only the PASAT2 (t(41) = 3.218, p = 0.003) and showed a tendency not reaching significance towards influencing WLG (t(41) = −1.948, p = 0.058). Men performed better than women in PASAT2 and women had better scores than men in WLG. Education was positively correlated with performance in SDMT and PASAT2. 3.3. Prevalence of cognitive dysfunction The overall prevalence of cognitive dysfunction was 47.5% when CIS patients were included, and 52.8% when CIS patients were excluded. Frequency of cognitive dysfunction observed for each group was: RR: 40.0%, PP: 56.5%, SP: 82.8%, and CIS: 27.3%. No control participant

103

was impaired in more than two measures. Table 3 shows the frequency of impairment in each measure for each group of patients. 3.4. Entire sample of MS patients Duration of disease correlated negatively with measures of verbal memory (SRT) and SDMT. EDSS correlated negatively with SRT measures, SDMT, and with PASAT3. Age was negatively correlated with performance in SRTL, SPARTi/d, and SDMT. The small impact of years of education on the performance of controls was not observed in patients. The BDI did not correlate with BRBN measures (Table 4). BDI scores and years of education were therefore not included as covariates in the analyses. Gender influenced differently performance of patients in the various tests compared to what we observed in the control group, as women performed better than men on SRT measures (SRTL: t(158) = − 2.346, p = 0.02; SRTC: t(158) = − 2.413, p = 0.02; SRTD: t(158) = − 3.394, p = 0.001), SDMT (t(158) = − 2.179, p = 0.03), and WLG (t(158) = − 3.248, p = 0.01). 3.5. Comparisons of performance between all groups Table 5 displays performance on all BRBN measures in RR, PP, SP MS patients, CIS and controls. The first multivariate analysis showed between-group differences in performance (F[36,768] = 3.034, p b 0,001, η2 = 0.125). The pairwise comparisons showed that the control group performed better than RR, PP, and SP group in all BRBN measures. The control group performed also better than patients with CIS in SPART measures, WLG, SDMT and PASAT, except in SRT measures. Large effect sizes were present when CIS patients were compared to control participants on SPART measures, WLG, and PASAT2. Medium effect sizes were present when the same groups were compared on SDMT and PASAT3. RRMS and PPMS patients had medium to large effects sizes for all the measures, as they performed poorer then control participants in all BRBN subtests. SPMS patients had large

Table 5 Mean scores (SD) of MS subgroups, CIS and controls on BRNBT tasks

SRTL SRTC SRTD SPARTi SPARTd SDMT PASAT3 PASAT2 WLG

RRMS

PPMS

SPMS

CIS

CTLS

38.75 (12.72) 24.75 (13.65) 7.20 (2.96) 19.96 (5.85) 7.12 (2.53) 48.24 (11.97) 39.15 (12.96) 27.61 (11.39) 21.33 (4.34)

31.52 (16.79) 18.87 (12.65) 6.48 (2.73) 17.96 (6.36) 6.26 (2.42) 37.30 (12.80) 32.00 (14.10) 24.12 (12.80) 20.26 (5.49)

29.55 (16.19) 17.48 (16.80) 5.17 (3.05) 15.69 (4.91) 5.48 (2.32) 34.00 (10.72) 32.14 (14.16) 23.97 (10.82 18.86 (4.58)

44.33 (15.16) 31.06 (16.87) 8.03 (2.89) 19.49 (5.14) 7.24 (1.48) 50.24 (12.47) 39.67 (16.31) 29.12 (13.94) 20.82 (4.10)

47.68 (10.95) 36.91 (14.11) 9.05 (2.24) 23.49 (4.59) 8.77 (1.48) 58.27 (8.99) 47.93 (9.85) 38.33 (9.79) 25.15 (4.36)

RRMS = relapsing–remitting; PPMS = primary progressive; SP = secondary progressive; CIS = clinically isolated syndrome; CLTS = controls; SRT = Selective Reminding Test: L = Long-Term Storage; C = Consistent Long-Term Storage; D = Total Delay Recall; SPART = 10/36 Spatial Recall: i = immediate recall; d = delayed recall; SDMT = Symbol Digit Modality Task; PASAT = Paced Auditory Serial Addition Task (3 = 3 s/digit; 2 = 2 s/digit); WLG = Word List Generation.

104

C. Potagas et al. / Journal of the Neurological Sciences 267 (2008) 100–106

Table 6 Effect sizes (Cohen d) for differences between controls and MS subgroups after controlling for age and gender (when the difference is significant (p b 0.05), the d is marked with an ⁎)

SRTL SRTC SRTD SPARTi SPARTd SDMT PASAT 3 PASAT 2 WLG

Control vs. RR

Control vs. PP

Control vs. SP

Control vs. CIS

0.87⁎ large 0.96⁎ large 0.57⁎ medium 0.78⁎ medium 0.87⁎ large 0.98⁎ large 0.79⁎ large 1.02⁎ large 0.92⁎ large

0.95⁎ large 1.15⁎ large 0.59⁎ medium 0.77⁎ medium 1.04⁎ large 1.67⁎ large 1.17⁎ large 1.10⁎ large 0.90⁎ large

1.14⁎ large 1.10⁎ large 0.91⁎ large 1.41⁎ large 1.56⁎ large 1.97⁎ large 1.16⁎ large 1.27⁎ large 1.31⁎ large

0.30 0.38 0.41 0.88⁎ large 1.11⁎ large 0.73⁎ medium 0.65⁎ medium 0.80⁎ large 0.99⁎ large

SRT = Selective Reminding Test: L = Long-Term Storage; C = Consistent Long-Term Storage; D = Total Delay Recall; SPART = 10/36 Spatial Recall: i = immediate recall; d = delayed recall; SDMT = Symbol Digit Modality Task; PASAT = Paced Auditory Serial Addition Task (3 = 3 s/digit; 2 = 2 s/digit); WLG = Word List Generation; Ctrl = controls; RR = relapsing–remitting; PP = primary progressive; SP = secondary progressive; CIS = clinically isolated syndrome.

effect sizes for all measures, when they were compared to control participants (Table 6). 3.6. Comparisons of performances between groups of patients

groups performed better than chronic-progressive MS patients (PPMS and SPMS) in the SDMT. When EDSS replaced age as covariate, no differences were found in performance between patients groups (F[27,387] = 0.923 ns). 4. Discussion

The second multivariate analysis including only patients also showed between-group differences (F[27,444] = 1.610, p = 0.029, η2 = 0.089). The pairwise comparisons showed that CIS patients performed better than patients with RRMS, PPMS and SPMS in SRTL and SRTC measures and the effect sizes were small for RRMS and medium for the chronicprogressive groups. CIS patients also performed better than chronic-progressive patients in SDMT, and the effect sizes were medium for PPMS and large for SPMS patients. RRMS patients performed better than SPMS on SDMT (large effect size) and on SRTD, SPARTi and SPARTd (small effect sizes). The differences between performances of RRMS and PPMS and between PPMS and SPMS on the other tests did not reach significance (Table 7). When the disease duration replaced age as covariate in the overall analyses the results slightly changed (F[27, 420] = 1.645, p = 0.023, η2 = 0.096). Patients with CIS performed better than PPMS patients in SRTL, and both CIS and RRMS

The present study confirmed for the first time in a Greek population the presence of cognitive deficits in all groups of patients with MS and also in patients with CIS. The overall prevalence of cognitive dysfunction in our patients was 52.8% with CIS patients excluded and 47.5% with CIS patients included, in accordance with the estimated prevalence of recent studies in ethnically different populations. [3,5,21]. The frequency of cognitive dysfunction declined progressively from SPMS to PPMS to RRMS and finally to CIS. Mean scores of the control group in BRBN measures were similar to the normative data obtained in different populations, [11,22,23] indicating that this battery is little influenced by language or cultural differences, thereby validating its use in different populations. However, age, years of education and gender seem to influence differentially the performance of each population in different tasks. Concerning the impact of

Table 7 Effect sizes (Cohen d) for differences between MS subgroups after controlling for age and gender (when the difference is significant (p b 0.05), the d is marked with an ⁎)

SRTL SRTC SRTD SPARTi SPARTd SDMT PASAT 3 PASAT 2 WLG

CIS vs. RR

CIS vs. PP

CIS vs. SP

RR vs. PP

RR vs. SP

PP vs. SP

0.45⁎ small 0.46⁎ small 0.48 − 0.03 0.08 0.2 0.02 0.09 0.02

0.60⁎ medium 0.62⁎ medium 0.36 b0.001 0.23 0.78⁎ medium 0.34 0.21 0.06

0.75⁎ medium 0.65⁎ medium 0.79⁎ medium 0.48 0.66⁎ medium 1.16⁎ large 0.36 0.27 0.37

0.22 0.16 b0.001 0.02 0.12 0.60 0.37 0.14 0.09

0.38 0.25 0.25⁎ 0.47⁎ 0.47⁎ 0.97⁎ 0.38 0.19 0.42

0.14 0.11 0.11 0.43 0.37 0.30 b0.001 0.01 0.28

small small small large

SRT = Selective Reminding Test: L = Long-Term Storage; C = Consistent Long-Term Storage; D = Total Delay Recall; SPART = 10/36 Spatial Recall: i = immediate recall; d = delayed recall; SDMT = Symbol Digit Modality Task; PASAT = Paced Auditory Serial Addition Task (3 = 3 s/digit; 2 = 2 s/digit); WLG = Word List Generation; RR = relapsing–remitting; PP = primary progressive; SP = secondary progressive; CIS = clinically isolated syndromes.

C. Potagas et al. / Journal of the Neurological Sciences 267 (2008) 100–106

gender, our results were very close to the normative data given by Sepulcre et al. [23] We found no significant differences in the performance of our control group with respect to gender, with the exception of PASAT2 in favor of men. The pattern of performance, however, was different in MS patients, favoring women on SRT, SDMT and WLG. Savettieri et al. [24] have demonstrated a possible gender-related effect of some clinical and genetic variables on cognitive impairment in MS, providing a plausible explanation for the differential impact of gender on MS cognitive dysfunction presently observed. The impact of years of education in this Greek sample was less significant than in the other populations (only present for SDMT and PASAT2 in the control group). This could be due to the relative homogeneity and small variability in years of education in both our control and patients' samples. Therefore, this variable was not introduced as a covariate in the multivariate analyses. All MS patients differed significantly from control participants in all BRBN measures. The only exception was CIS patients, whose performance in verbal learning/ memory measures (SRT) did not differ significantly from controls. It has been suggested that the neuropsychological batteries specifically devised to evaluate cognitive disorders of MS patients (including the BRBN) are not very sensitive in the early stages of the disease and so are not optimal predictors of functional impairment in MS. [25] Our results support the BRBN as not only a brief, but also a sensitive measure of cognitive impairment in MS from the earliest stages. We could not reveal two distinct cognitive profiles based on the BRBN, as has been reported in a recent study [3]. Huijbregts et al. conclude that RRMS and SPMS patients have poor spatial working memory (SPART) and semantic fluency (WLG), relative to PPMS patients. In our patient sample, PPMS patients did not perform significantly different on SPART and WLG tasks from other MS subtypes. Our PPMS group seems to have a greater incidence of cognitive impairment than previously reported [26]. Adopting a close approximation of the Camp et al. criteria, the incidence of cognitive impairment in PPMS was 56.5% compared with the 28.6% previously reported. This difference might be partially explained by a chance effect due to the small number of PP patients in the present study (23 instead of 157). It is noteworthy that in the present study the disease duration of PP patients is shorter compared to the previous study [26] (mean years of 4.7 ± 5.3 vs. 10.9 ± 7.0). PP patients in Camp's study performed significantly worse than the controls in tests of verbal memory, attention and verbal fluency, in accordance with our results, but not in SPART as our PPMS group. Camp et al. hypothesized that PP patients were not deficient on the spatial memory task because of the low ceiling level of the SPART. An important question that has arisen in the literature is whether the pattern of cognitive impairment observed in MS patients can be explained primarily in terms of information processing speed dysfunction [27,28]. Although the tasks of

105

BRBN are not easily distinguished along these lines [3], as multiple cognitive mechanisms may be involved in each one of them, an attempt was made to address this issue. Our results, however, were inconclusive. The SRT is the task with the lowest speed processing demands, being primarily a task of short and long-term memory. We found that RRMS, SPMS and PPMS patients were significantly impaired on the SRT compared to controls, with large Cohen d effect sizes, whereas CIS patients were not significantly impaired on SRT measures compared to the control group. This could suggest that tasks with higher processing speed demands are more susceptible in the early stages of the disease. However, CIS patients were most frequently impaired in semantic verbal fluency (WLG: 33.3%), followed by spatial delayed recall (SPARTd: 27.3%), the same measures being the more frequently impaired in Achiron and Barak's study. [4] Different cognitive (working memory) processes could be involved in those tasks, but the decline cannot be explained solely by an impaired speed processing mechanism (for the WLG we limit this effect by giving 90 s time in MS patients and for SPARTd there is no time limit). On the other hand, CIS patients showed the greatest discrepancy from chronicprogressive types of MS on the SDMT, a task with significant speed processing demands, requiring a working memory activity and quick communication across different modalities. RRMS patients were also significantly less impaired compared to patients with SPMS on the SDMT, in accordance with a previous cross-sectional study [2]. The SDMT showed, therefore, a relative progression of the processing speed slowing in MS, accompanying disease progression. However, this was not observed on the PASAT, with RRMS and CIS patients not significantly less impaired than SPMS patients. The PASAT is also used in BRBN as a measure of processing speed and working memory along with the SDMT. One possible explanation for this discrepancy might be that the nature of the working memory abilities differs between the BRBN tasks. [3] Beside higherlevel integrative deficits, covert sensory-motor deficits could also influence performance on these tasks. A sub-vocal articulatory loop could intervene with the execution of the PASAT, whereas oculo-motor disturbances and slowed down visual scanning and tracking with the execution of the oral version of SDMT. [25] In the present study, CIS patients were less frequently impaired on different BRBN measures, than suggested in a previous study [4]. This is probably due to the different criterion used for considering patients as impaired in a task. In the above study a patient was considered as impaired at only 1 SD below the mean for normal controls, whereas in our study impaired patients scored below the 5th percentile for normal controls. Moreover, the EDSS score in our CIS patients was lower and the disease duration longer. Concerning RRMS patients, disease duration of RRMS patients was longer than in the previous study, but mean EDSS score was lower. Differences between progressive forms and RRMS and CIS patients were observed after controlling for age, but

106

C. Potagas et al. / Journal of the Neurological Sciences 267 (2008) 100–106

largely disappeared after controlling for physical disability (EDSS), suggesting that disease severity accounts for more variation in cognitive performance than age as previously observed. [2] EDSS was the only variable that correlated significantly with the BDI score. However, BDI score showed no correlation with cognitive impairment, and was, therefore, excluded as a covariate. In conclusion, findings from the present study confirm for the first time in the Greek population that cognitive dysfunction should be considered among the main manifestations of MS even in the early stages of the disease. The BRBN appears sensitive in detecting neuropsychological impairment in all MS subtypes, including CIS. Significant impairment occurs across all studied cognitive domains in all MS subtypes failing to detect distinct patterns of impairment between MS subtypes. Findings suggest a relatively global pattern of cognitive deficits in MS independently of the disease course, but with gradual augmentation of the frequency of appearance depending on the progression of the disease. The pattern of cognitive dysfunction in patients with CIS suggestive of MS is similar, although verbal learning and memory capacity remain relatively spared. References [1] Achiron A, Barak Y. Cognitive changes in early MS: a call for common framework. J Neurol Sci 2006;245:47–55. [2] Huijbregts SCJ, Kalkers NF, de Sonneville LMJ, de Groot V, Reuling IEW, Polman CH. Differences in cognitive impairment of relapsing–remitting, secondary and primary progressive MS. Neurology 2004;63:335–9. [3] Huijbregts SCJ, Kalkers NF, de Sonneville LMJ, de Groot V, Polman CH. Cognitive impairment and decline in different MS subtypes. J Neurol Sci 2006;245:187–94. [4] Achiron A, Barak Y. Cognitive impairment in probable MS sclerosis. J Neurol Neurosurg Psychiatry 2003;74:443–6. [5] Bobholz JA, Rao SM. Cognitive dysfunction in multiple sclerosis: a review of recent developments. Curr Opin Neurol 2003;16(3):283–8. [6] Rao SM, Leo GJ, Bernardin L, Unverzagt F. Cognitive dysfunction in multiple sclerosis. I. Frequency, patterns and prediction. Neurology 1991;41:685–91. [7] Zakzanis KK. Distinct neurocognitive profiles in multiple sclerosis subtypes. Arch Clin Neuropsychol 2000;15(2):115–36. [8] Foong J, Rosewicz L, Chong WK, Thompson AJ, Miller DH, Ron MA. A comparison of neuropsychological deficits in primary and secondary progressive multiple sclerosis. J Neurol 2000;247:97–101. [9] Comi G, Filippi M, Martinelli V, Camo A, Rodegher M, Alberoni M, et al. Brain MRI correlates of cognitive impairment in primary and secondary progressive multiple sclerosis. J Neurol Sci 1995;132:222–7. [10] Gaudino EA, Chiaravalloti ND, DeLuca J, Diamond B. A comparison of memory performance in relapsing–remitting, primary progressive,

[11]

[12]

[13] [14] [15] [16] [17] [18]

[19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

and secondary progressive, multiple sclerosis. Neuropsychiatry Neuropsychol Behav Neurol 2001;14(1):32–44. Boringa JB, Lazeron RH, Reuling IE, Ader HJ, Pfennings L, Lindeboom J, et al. The brief repeatable battery of neuropsychological tests: normative values allow application in multiple sclerosis clinical practice. Mult Scler 2001;7:263–7. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an Expanded Disability Status Scale (EDSS). Neurology 1983;33:1444–52. Beck AT, Ward CH, Medelson M, Mock J, Erlbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71. Buschke H, Fuld PA. Evaluating storage, retention, and retrieval in disordered memory and learning. Neurology 1974;24:1019–25. Smith A. Symbol Digit modalities test. Los Angeles: Western Psychological Services; 1991. Gronwall DM. Paced auditory serial-addition task: a measure of recovery from concussion. Percept Mot Skills 1977;44:367–73. Basso MR, Beason-Hazen S, Lynn J, Rammohan K, Bornstein RA. Screening for cognitive dysfunction in multiple sclerosis. Arch Neurol 1996;53:980–4. Denney DR, Sworowski LA, Lynch SG. Cognitive impairment in three subtypes of multiple sclerosis. Arch Clin Neuropsychology 2005;20:967–81. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale: Lawrence Earlbaum Associates; 1988. Amato MP, Zipoli V, Portaccio E. Multiple sclerosis-related cognitive changes: a review of cross-sectional and longitudinal studies. J Neurol Sci 2006;245:41–6. Amato MP, Portaccio E, Goretti B, Zipoli V, Ricchiuti L, DeCaro MF, et al. The Rao's Brief Repeatable Battery and Stroop test: normative values with age, education and gender corrections in an Italian population. Mult Scler 2006;12:787–93. Sepulcre J, Vanotti S, Hernandez R, Sandoval G, Caceres F, Garcea O, et al. Cognitive impairment in patients with multiple sclerosis using the Brief Repeatable Battery-Neuropsychology test. Mult Scler 2006;12:187–95. Savettieri G, Messina D, Andreoli V, Bonavita S, Caltagirone C, Cittadella R. Gender-related effect of clinical and genetic variables on the cognitive impairment in multiple sclerosis. J Neurol 2004;251:1208–14. Gainotti G. Measures of cognitive and emotional changes in multiple sclerosis and underlying models of brain dysfunction. J Neurol Sci 2006;245:15–20. Camp SJ, Stevenson VL, Thompson AJ, Miller DH, Borras C, Auriacombe S, et al. Cognitive function in primary progressive and transitional progressive multiple sclerosis. A controlled study with MRI correlates. Brain 1999;122:1341–8. De Sonneville LMJ, Boringa JB, Reuling IEW, Lazeron RHC, Ader HJ, Polman CH. Information processing characteristics in subtypes of multiple sclerosis. Neuropsychologia 2002;40:1751–5. Demaree HA, De Luca J, Gaudino EA, Diamond B. Speed of information processing as key deficit in multiple sclerosis: implications for rehabilitation. J Neurol Neurosurg, Psychiatry 1999;67:661–3.